For AI agents: the complete documentation index is available at https://open-spaced-repetition.github.io/ts-fsrs/llms.txt, the full documentation bundle is available at https://open-spaced-repetition.github.io/ts-fsrs/llms-full.txt, and this page is available as Markdown at https://open-spaced-repetition.github.io/ts-fsrs/guide/model/fsrs-6.md.

FSRS-6

FSRS-6 has 21 trainable weights, including a trainable forgetting-curve decay and a same-day stability term.

See the Model overview for the shared symbols and Model contract.

Formula source

Reviewed against awesome-fsrs Wiki — FSRS-6.

Forgetting curve and interval

With the positive decay magnitude d=w20d = w_{20}, choose FF so that R(S,S)=0.9R(S, S) = 0.9:

F=0.9−1/d−1,R(t,S)=(1+FtS)−dF=0.9^{-1/d}-1,\qquad
R(t,S)=\left(1+F\frac{t}{S}\right)^{-d}
I(r,S)=SF(r−1/d−1)I(r,S)=\frac{S}{F}\left(r^{-1/d}-1\right)

nextInterval rounds the result to whole days and returns at least one day.

Initial memory

S0(G)=clamp⁡(wG−1,Smin,Smax)S_0(G)=\operatorname{clamp}\left(w_{G-1},S_{min},S_{max}\right)
D0(G)=w4−ew5(G−1)+1,Dinitial=clamp⁡(D0(G),1,10)D_0(G)=w_4-e^{w_5(G-1)}+1,\qquad
D_{initial}=\operatorname{clamp}\left(D_0(G),1,10\right)

Difficulty

ΔD=−w6(G−3),D∗=D+ΔD10−D9\Delta D=-w_6(G-3),\qquad
D^*=D+\Delta D\frac{10-D}{9}
D′=clamp⁡(w7D0(4)+(1−w7)D∗,1,10)D'=\operatorname{clamp}\left(w_7D_0(4)+(1-w_7)D^*,1,10\right)

Stability after review

For t>0t > 0, or when enableShortTerm is disabled, the rating selects the following branch.

The update uses the pre-review DD and SS. HG=w15H_G = w_{15} for Hard and 11 otherwise; EG=w16E_G = w_{16} for Easy and 11 otherwise.

When enableShortTerm is enabled, CC is the previous stability multiplied by exp⁡(−w17w18)\exp(-w_{17}w_{18}); otherwise, C=SC = S:

Sf=min⁡(w11D−w12((S+1)w13−1)ew14(1−R),C)S_f=\min\left(w_{11}D^{-w_{12}}\left((S+1)^{w_{13}}-1\right)e^{w_{14}(1-R)},C\right)
Sr=S[1+ew8(11−D)S−w9(ew10(1−R)−1)HGEG]S_r=S\left[1+e^{w_8}(11-D)S^{-w_9}\left(e^{w_{10}(1-R)}-1\right)H_GE_G\right]
Sraw′(G)={Sf,G=1Sr,G∈{2,3,4}S'_{raw}(G)=
\begin{cases}
S_f,&G=1\\
S_r,&G\in\{2,3,4\}
\end{cases}

CC is an implementation cap for post-lapse stability, not a separately configured stability floor. The model then clamps Sraw′S'_{\mathrm{raw}} to the range from Smin⁡S_{\min} to Smax⁡S_{\max}.

Same-day update

Here, “same-day” means the model receives elapsedDays = 0; the model does not inspect calendar dates. Direct callers and the scheduler’s chrono determine the elapsed value.

When enableShortTerm is enabled and t=0t = 0:

sinc=S−w19ew17(G−3+w18)s_{inc}=S^{-w_{19}}e^{w_{17}(G-3+w_{18})}
S′=clamp⁡(Sm,Smin,Smax),m={sinc,G=1max⁡(sinc,1),G≥2S'=\operatorname{clamp}\left(Sm,S_{min},S_{max}\right),
\qquad
m=\begin{cases}
s_{inc},&G=1\\
\max(s_{inc},1),&G\ge2
\end{cases}

Hard, Good, and Easy cannot reduce stability below its current value; Again may reduce it.